Hello,

TLDR: I recently built a water cooling loop with several AquaComputer devices, and having any of them connected via USB prevents the system from going to sleep, keeping the displays active. Since I have OLED monitors, this is particularly problematic, as I am worried about display burn-in.

Devices I have:

• LeakdShield
• Octo (w/ 6 fans connected)
• Ultitube D5 Next pump/reservoir combo
• High Flow Next
• Hubby USB hub

Things I have tried:

• Connecting the LeakShield directly to the motherboard USB header (displays stay on)
• Connecting all the devices to the Hubby hub one at a time with the Hubby set to USB power mode (displays stay on if one or more AquaComputer devices are connected)
• Connecting all the devices to the Hubby hub one at a time with the Hubby set to SATA power mode and a SATA cable connected (displays stay on if one or more AquaComputer devices are connected)
• Connecting the Hubby only but no devices connected to the Hubby (this allows the displays to turn off)
• Updating firmware and driver on all of my devices- AquaComputer, AquaSuite, system, etc.
• Running "powercfg -requests" in a command line to see what is keeping the system awake (no results shown- all areas indicate "None")
• Running "powercfg /devicequery wake_armed" in a command line to see what devices are able to wake the system/keep the system awake (see output below)

My motherboard is an ASUS ROG ZENITH EXTREME with an X399 chipset and an AMD ThreadRipper 1950X CPU. When I was first setting up the AquaComputer devices, I tried connecting them to a NZXT hub which was connected to the motherboard USB headers. However, I got an error/warning from the AquaSuite setup software saying it could not detect the devices and that it was a known issue with X399 chipsets (and several others). The computer detected the devices (they were present in Device Manager), and AquaSuite itself detected they were present, it just could not read the serial numbers to perform registration. After much trial and error, I was finally able to get AquaSuite to detect the devices AND read the serial numbers, by connecting the devices to the AquaComputer Hubby and connecting that hub to an external USB port on my motherboard.

One theory that I have is that the AquaComputer devices are pretending to be mice/keyboard devices, as I only have one of each connected, but the results of the powercfg command show multiples. I have also looked in Device Manager and noted that with the AquaComputer devices all connected via the Hubby, I have 5 HID Keyboard Devices listed. However, if I disconnect the USB hub, I only have 4 HID Keyboard Devices listed. When the Hubby and all 4 AquaComputer devices are connected, I also have one additional "HID-compliant consumer control device", one additional "HID-compliant system controller", an extra 4 "HID-compliant vendor-defined device", and an extra 5 "USB Input Device".

My only question:
How do I allow the system to go to sleep/displays to turn off, without having to disconnect the AquaComputer devices?


Output from "powercfg /devicequery wake_armed" command:
HID-compliant mouse
Xbox Gaming Device
HID-compliant mouse (001)
HID Keyboard Device (002)
HID-compliant mouse (002)
HID Keyboard Device (005)

After a bit more searching, I found these posts:
Aqua devices preventing monitor/system sleep?
OCTO prevents monitor timeout and sleep?

Which references this Reddit thread:
https://www.reddit.com/r/watercooling/co…web2x&context=3

Long story short, it appears that the Alienware Command Center somehow interacts with the AquaComputer services to keep the system awake. While I did not have the same Alienware services running that were referenced in the threads above (and in fact, I could not find any Alienware "Services" running in Task Manager), going to the "Details" tab in Task Manager did show a few Alienware processes running- the AlienFXSubAgent and a few AWCC processes. Killing the AlienFXSubAgent process would just spawn a new one. However, after uninstalling the Alienware Command Center, my system now goes to sleep as it should, so my problem appears to be solved.